2048
J. Am. Chem. SOC.1982, 104, 2048-2050
characterization of 19, and mono- or dibenzyl derivatives have been prepared and characterized from 20a,b.16 These results show that the greater stability of ester relative to thiol ester is sufficient to dominate over ring size effects. In the most demanding 6- to 8-membered ring conversion 17 19, the result is somewhat obscured by competing dimer formation, but the trends are clear. Differences in ring strain between thiol lactone and mercapto lactone isomers are only important in the reactions 17 19 and 16 18. In larger rings, the lactone is favored by a clear margin. Synthetically useful conversions of hydroxyalkyl thiol lactones to mercapto lactones are expected in the absence of drastic changes in strain or transannular interactions.
-
-
-
Acknowledgment. This work was supported by the National Science Foundation (Grant CHE-8113026). Registry No. la, 79815-79-5; lb, 79815-88-6; 3, 81044-80-6; 3a, 81044-81-7; 3b, 81044-82-8; 4a, 81044-83-9; 4b, 81098-22-8; Sa, 81044-84-0; Sb, 81044-85-1; 721, 81044-86-2; 7b, 81044-87-3; 8a, 81044-88-4; 8b, 81064-08-6; 9a, 81044-89-5; 9b, 81044-90-8; 10, 81044-91-9; 11, 61448-27-9; 13, 81044-92-0; 14, 81044-93-1; 15, 81044-94-2; 16, 8 1044-95-3; 17, 8 1044-96-4; 18, 81044-97-5; 19, 81044-98-6; ZOa, 81044-99-7; ZOb, 81045-00-3; 21, 81045-01-4; 2-diphenylphosphinyl-7-(dimethylbutyl)silyloxymethylthiepane,8 1064-09-7; 2-diphenylphosphinyl-6-(2-hydroxy)ethyl-tetrahydrothiopyran,8104502-5. (15) 21 (oil): NMR (270 MHz, CDClS) 6 7.3 ( 5 H, m), 4.3 (1 H, ddd, J = 11.4, 5.9, 4.0 Hz), 4.1 (1 H, ddd, J = 11.4, 5.5, 4.0 Hz), 3.7 (2 H, s), 2.57 (1 H, m), 2.15 (2 H, two overlapping dt, J = 15.4,6.6 Hz and J = 15.4, 6.3 Hz), 1.8 (2 H, JAB= 5.5 Hz), 1.67 (2 H, m), 1.48 (4 H, m); IR (neat) 1740 cm-I. (16) Satisfactory exact mass data have been obtained for all lactones and dilactones. (17) A solution of 2-lithiothiepane S-oxide was added to e x m s dimethyl carbonate in THF (at 20 "C, 45% yield of 2-carbomethoxythiepaneS-oxide). Reduction with NaBH, in ethanol-THF (1.5 h, 20 "C) gave 14 (82%). (18) Cer6, V.; Paolucci, C.; Pollicino, S.;Sandri, E.; Fava, A. J . Org. Chem. 1978,43, 4826.
Suicidal Inactivation of Cytochrome P-450 by Cyclopropylamines. Evidence for Cation-Radical Intermediates Robert P. Hanzlik* and Robert H. Tullman Department of Medicinal Chemistry University of Kansas, Lawrence, Kansas 66045 Received November 17, 1981 Cyclopropylamine derivatives are known to have interesting and sometimes useful properties as enzyme inhibitors. For example Coprine (l), a constituent of inky-cap mushrooms (Coprinus sp.),
9
1
6R=CH,
is hydrolyzed in vivo to 1-hydroxycyclopropylamine and cyclopropanone hydrate, which inhibit aldehyde dehydrogenase.'J This in turn gives rise to a disulfiram-like reaction if these mushrooms are ingested with ethanol. N-Benzylcyclopropylamine(BCA, 2)3 and related arylalkyl cycl~propylamines~ are inhibitors of mito(1) Tottmar, 0.; Lindberg, P. Acta Pharmacol. Toxicol. 1977, 40, 476-48 1. (2) Wiseman, J. S.;Abeles, R. H. Biochemistry 1979,18, 427-435. (3) Silverman, R. B.; Hoffman, S . J. J . Am. Chem. SOC. 1980,102,884.
Table 1. Covalent Binding of Radioactivity to Microsomal Proteinsa nmol bound
incubation conditions
time, min
expt A, [7-3H,7-14C]-2 standard -NADPH standard -NADPH
10 10 60 60
2.49 0.20 6.63 0.16
* 0.09 * 0.05 * 1.24 * 0.03
2.03 i 0.50 0.10 + 0.05 6.54 i 0.89 0.22 i 0.04
expt B, [7-'H,7-"T]-2 standard +glutathione (1 mM) +semicarbazide (0.1%)
60 60 60
3.74 0.92 2.56
* 0.27 * 0.05 * 0.13
3.86 t 0.40 1.11 i 0.17 2.60 i 0.14
expt C, [ 7-'H] -6 standard +glutathione (1 mM)
60 60
0.88 0.48
'H
c
"
Standard conditions were as described in ref 6. Each experiment used a different preparation of microsomes, but these results are typical of several such experiments. At the indicated times aliquots of incubation mixture were withdrawn and covalently bound radioactivity was measured by using method A (ref 9). Results are net binding after correction for a small zero-time background and are expressed as the mean +SD (n = 3) or averages of duplicates.
chondrial monoamine oxidase (MAO, E.C. 1.4.3.4) and tranylcypromine (3) is a therapeutically useful M A 0 inhibitor. We recently reported that 2 and a number of its derivatives were potent ~ characteristics inhibitors of cytochrome P-450 e n ~ y m e s . Several of the inhibition process suggested that it might involve suicide inactivation of the enzyme via a metabolite of the parent amine. Thus, loss of enzyme activity followed first-order kinetics, required oxygen and NADPH, and was inhibited by carbon monoxide but not by glutathione. In this corhmunication we report further studies that firmly establish the suicidal nature of the enzyme inactivation process and suggest the involvement of a novel mechanism for the enzymatic activation of cyclopropylamines by cytochrome P-450. Under standard conditions,6 incubation of 2 with rat liver microsomes leads to a first-order loss of aminopyrine demethylase activity with a half-life of 14.9 min,' as shown in Figure la. So that it could be determined whether the inactivation of P-450 by 2 involved covalent modification of the enzyme, [7-3H]-2was prepared* and incubated with microsomes under the standard (4) (a) Murphy, D. L.; Donnelly, C. H.; Richelson, E.; Fuller, R. W. Biochem. Pharmacol. 1978,27, 1767. (b) Fuller, R. W.; Hemrich, D. K.; Mills, J. Ibid. 1978,27,2255. (c) Long, R. F.; Mantle, T. J.; Wilson, K. Ibid. 1976,25,247. (d) Winn, M.; Horrom, B. W.; Rasmussen, R. R.; Chappel, R. B.; Plotnikoff, N. P. J. Med. Chem. 1975,18, 437. (5) Hanzlik, R. P.; Kishore, V.; Tullman, R. J . Med. Chem. 1979,22,759. (6) Liver microsomes were prepared from male rats (2OC-300 g) and washed by resuspension in 1.15% KCI and recentrifugation at 105000g. Incubations contained 100 mg of microsomes (ca. 8 mg of protein)/mL in 0.1 M NaK phosphate buffer (pH 7.6) containing 1 mM EDTA, 7 mM MgCI2, 6.6 mM glucose-6-phosphate, 0.65 mM NADP, and 1-2 IU of glucose 6phosphate ddehydrogenase/mL. Incubations were carried out at 33 OC under air. Aminopyrine was added at 3.5 mM (along with 0.1% semicarbazide)and incubated for 5 min; formaldehyde was measured by the Nash procedure. Cyclopropylamines2 and 6 were added to incubations to a concentration of 1 mM. (7) It is probable that the total inhibition observed at a given time includes a reversible component as well as that due to enzyme inactivation by 2 or 6 and that significant inactivation of cytochrome P-450 by 2 or 6 occurs even during the 5-min assay with aminopyrine. Hence the exact amount of enzyme remaining at a given time is somewhat ambiguous, and inactivation plots tend to be. shifted to the right of the covalent binding plots. However, neither of these eventualities interfere with measurement of the rate of change in enzyme activity with time and the associated half-life for inactivation under standard conditions. (8) Amines 2 and 6,tritiated on the benzylic carbon, were prepared by reduction of the corresponding benzylidene Schiff bases in absolute ethanol with [)H]NaBH4. For incubations they were diluted with cold carrier to final specific activities of 2-6 Ci/mol. Carbon-14 labeled 2 was prepared from [~arboxyl-'~C] benzoic acid by successive treatment with oxalyl chloride in benzene, excess cyclopropylamine in ether, and excess BHytetrahydrofuran at 25 OC for 24 h. For incubations [I4C]-2was used at a specific activity of 0.7 Ci/mol.
0002-7863/82/ 1504-2048$01.25/0 0 1982 American Chemical Society
J . Am. Chem. SOC.,Vol. 104, No. 7. 1982 2049
Communications to the Editor
80
$
0.2
=
60
I I )' a
20
A 1 I
I
4
8
I
12 MINUTES
I
I
16
20
possible involvement of benzylic activation in the covalent binding process could be ruled out, [7-14C,7-3H]-2was prepareds and incubated with microsomes. As shown in Table I, the 3H/'4C ratio in the bound material was the same as that of the starting material. Addition of glutathione (GSH) was found to depress the total amount of covalent binding, but it did not affect the 3H/'4C ratio of the bound material. In contrast, GSH had no effect on either the rate of extent of loss of enzyme activity. This result suggests that covalent incorporation of 2 specifically involves the cyclopropyl-nitrogen moiety, a structural feature known to be an absolute requirement for enzyme i n a c t i v a t i ~ n . ~ J It ~ Jalso ~ suggests that the reactive species generated by the enzyme can diffuse from the active site and react either with glutathione or other protein groups in reactions that do not lead to enzyme inactivation. Presumably, therefore, enzyme inactivation results from the covalent binding of a reactive metabolite of 2 to parts of the enzyme active site. In the case of compound 2 the usual a-hydroxylation mechanism for N-dealkylation of amines by cytochrome P-450 would lead to carbinolamine 4. Such adducts of cyclopropanone readily OH
PhCH2-NH
I
IO
I
20
30
40
!io
MINUTES
Figure 1. Inactivation of P-450 (A) and covalent modification of microsomal protein ((0)B = dpm/mg protein) by N-benzylcyclopropylamine (part A) and N-benzyl-1-methylcyclopropylamine (part B). The incubation and assay conditions are given in ref 6 and 9.
conditions for the inactivation reaction. As shown in Figure l a , tritium became covalently bound9 to hemefree microsomal protein, also with first-order kinetics and a half-life of 14.9 min, suggesting that both binding and enzyme inactivation involve a common enzymatically generated reactive intermediate." So that the (9) Covalent binding of radioactivity to microsomal protein was determined by one of two methods. Method A involved treating the microsomal suspensions with 4 volumes of MeOH/CHCI3 (4:l) followed by repetitive washing of the precipitate with MeOH/H20 (4:l)and fmally redissolving the precipitate in 1 N NaOH for counting. This method was shown to remoue all detectable heme from the precipitated proteins. Method B involved treating the microsomal suspensions with 10 volumes of MeOH/HSO4 (95:5) overnight at 4 OC in the dark, followed by centrifugation and washing of the pellet with MeOH/H20 (4:l). The pellet was dissolved in 1 N NaOH for counting as above. The original MeOH/H2S04 supernatant was then processed for analysis of covalently modified heme as described by Ortiz de Montellano and Mico.'O Preliminary studies with ['H]-2 have repeatedly indicated that tritium label becomes covalently bound to both heme and protein. Over several experiments with and without glutathione the ratio of heme to protein binding varied from 2:3 to 3:2,but both portions were consistently labeled. Most of the heme-associated tritium was associated with a band of 'abnormal green pigments" having an R,of 0.82 in CHC13/acetone (3:l)on silica G plates. The nature of the bound material is currently under investigation. (10) Ortiz de Montellano, P.R.; Mico, B.A. Molec. Pharmacol. 1980,18, 128-135. ( 1 1) Although it is not apparent from the experiments shown in Figure 1, inactivation by 2 (or 6 ) of aminopyrine-N-demethylasein liver microsomes from uninduced rats plateaus after loss of only 70-80% of the original activity, whereas covalent binding continues for a bit longer time before it too plateaus (data not shown). This pattern of activity is consistent with the well-known existence in microsomes of multiple forms of cytochrome P-450having different but partly overlapping specificities for substrates, inhibitors, and inducers. Thus apparently some demethylases active on aminopyrine are not affected by 2 (or 6 ) while others not active on aminopyrine may be capable of activating 2 (and 6 ) to species that bind to microsomal proteins; the apparent overall half-lives for the two processes are very similar.
PhCH2-63(l H
exchange ligands at C-1 via intermediates such as 5 or related isoelectronic structures in which 0 or S replaces the nitrogen atom,I4 and 5 has been proposed3 as the active metabolite responsible for the suicide inactivation of M A 0 by 2. According to this view, N-benzyl-1-methylcyclopropylamine (6), which cannot form a cyclopropylidene Schiff base analogous to 5, should not be an inactivator of cytochrome P-450. We were thus surprised to find that 6 is indeed a very effective inactivator of cytochrome P-450. The inactivation process requires oxygen and NADPH (data not shown) and under standard conditions6follows first-order kinetics with a half-life of 12.1 min (Figure lb). Furthermore, as with 2, tritium from [7-3H]-6 is covalently incorporated into microsomal protein as the inactivation proceeds (tl/z = 10.7 min), and only a portion of this incorporation can be blocked by glutathione (Table I). For comparison, N-benzyl-tert-butylamine (7) was tested and found to be a very weak time-independent inhibitor of cytochrome P-450, much like N-benzylisopropylamine.5 Thus, although 6 may not be capable of forming a cyclopropylidene Schiff base (e.g., 5), it is clearly a suicide substrate for cytochrome P-450 and its activity also is associated with the cyclopropyl-nitrogen moiety. An intriguing aspect of cytochrome P-450 action is its ability to utilize organic peroxides rather than NADPH and oxygen to carry out the oxidation of many of its substrates. Although the intermediates involved and the extent of similarity in mechanism between the P-450 peroxidase and mixed-function oxygenase processes are not clear, the fact that cytochrome P-450 catalyzes the cumene hydroperoxide dependent N-demethylation of aminopyrine via an aminium cation-radical intermediateI5J6 suggests a plausible mechanism for the activation of 6 (and possibly of 2). Thus, one-electron oxidation of 6 to an aminium ion such as 8b could activate the adjacent and already cyclopropyl ring for covalent reaction with the enzyme. Several modes of reaction are conceivable, as suggested in Scheme I. Path a would involve ring expansion and loss of a proton and an electron, leaving a quaternized azetidinium ion which could undergo nucleophilic attack (12)Tullman, R. H.; Hanzlik, R. P.Deu. Biochem. 1979, 6, 187. (13)Hanzlik, R.P.;Tullman, R.H.; Kishore, V. In "Microsomes, Drug Oxidations, and Chemical Carcinogenesis"; Coon, M. J., Conney, A. H., Estabrook, R. W., Gelboin, H. V., Gillette, J. R., OBrien, P.J., a s . ; Academic Press: New York, 1980;Vol. I, pp 327-30. (14)For examples and further references see: Jorritsma, R.; Steinberg, H.; deBoer, Th. J. Recl. Trau. Chim. Pays-Bas 1981,100, 184-194,194-200. (15)Ashley, P.L.; Griffin, B. W. Molec. Pharmacol. 1981, 19, 146-152. (16)Griffen, B. W.;Marth, C.; Yasukochi, Y.; Masters, B. S . S. Arch. Biochem. Biophys. 1980, 205, 543-553.
2050
J. Am. Chem. SOC.1982, 104, 2050-2052
Scheme I
n o/ a
Ph
Mechanism of Cytochrome P-450 Inhibition by Cyclopropylamines Timothy L. Macdonald,*l Karimullah Zirvi,2 Leo T. Burka, Pamela Peyman, and F. Peter G ~ e n g e r i c h * ~
6h
9
Departments of Chemistry and Biochemistry and Center in Environmental Toxicology Vanderbilt University, Nashville, Tennessee 37235
R
8a, R = H 8b, R = CH,
Received October 5, 1981 10
by a group on the enzyme. Precedent for this mode of reactivity is found in the oxidative rearrangement of N-monosubstituted 1-aminocyclopropanols to @-lactamswith t-BuOC1 followed by silver ion." Path b would simply involve opening of the cyclopropyl ring and reaction with a group from the enzyme. However, while this mode of reaction is well precedent4 in the reaction of cyclopropyl alcohols18 and thiols14 by one-electron oxidants, it is doubtful that an adduct such as 10 would be sufficiently stable toward hydrolysis to account for the covalent binding of benzylic 3H and I4C, which is observed even after harsh acid treatmentg of the microsomal proteins. While 4 remains a potential intermediate in the metabolic activation of 2 by cytochrome P-450, it is also possible that both 2 and 6 are activated via aminium radical cations as suggested in Scheme I. Several additional observations are consistent with this hypothesis. First, the cyclopropyl moiety is resistant to attack by extremely strong oxidants such as RuO4I9or dry which readily oxidize aromatic and aliphatic positions adjacent to a cyclopropane ring. Second, cyclopropyl ethers and thioethers are apparently not inactivators of P-45Ol2 despite their potential for metabolism to reactive cyclopropanone hemiketals or hemithioketals analogous to 4. However, the contrasting behavior of ethers and thioethers to that of amines 2 and 6 may be understandable in terms of cation-radical intermediates. Oxygen cation radicals are much less likely to form because of the high ionization potential of ethers, and sulfur cation radicals, if formed,21,2zmay react further to give sulfoxide metabolites rather than enzyme inactivation. Finally, the very recent report that 6 is a suicide substrate for monoamine which we had independently lends further credence to the possible involvement of aminium cation radicals in the metabolic activation of cyclopropyl amines and possibly in the oxidative metabolism of amines in general. Studies directed toward trapping and identifying of enzymatically activated forms of 2 and 6 are presently under way in our laboratory. Acknowledgment. We thank the N I H for financial support of this work through a predoctoral traineeship (GM-07775) to R.H.T. and Grant GM-22357. Registry No. 2, 13324-66-8; [7-3H]-2, 80926-12-1; [I4C]-2,8092613-2; 6, 79432-92-1; [7-"]-6, 80926-14-3; 7, 3378-72-1; cytochrome P-450, 9035-5 1-2; N-benzylidenecyclopropylamine, 3 187-77-7; N benzylidene( 1-methylcyclopropyl)amine, 80926-1 5-4; [carboxyl-"C]benzoic acid, 1589-66-8; cyclopropylamine, 765-30-0. (17) Wasserman, H. H.; Adickes, H. W.; deOchoa, 0. E. J . Am. Chem. SOC.1971, 93, 5586. (18) Schaafsma, S.E.; Molenaar, E. J. F.; Steinberg, H.; deBoer, Th. J. Rec. Trav. Chim. Puys-Bus 1967, 86, 1301. (A masthead printing error lists this article as belonging to volume 87 of 1968). (19) Hanzlik, R. P., unpublished observations. (20) Proksch, E.; deMeijere, A. Angew. Chem., Int. Ed. Engl. 1976, 15, 761. (21) May, S.W.; Phillips, R. S.;Mueller, P. W.; Herman, H. H. J. B i d . Chem. 1981, 256, 8470. (22) Watanabe, Y.; Numata, T.; Iyanagi, T.; Oae, S.Bull. Chem. SOC. Jpn. 1981, 54, 1163. (23) Silverman, R. B.; Hoffman, S. J. Biochem. Biophys. Res. Commun. 1981. ..., 101. . 1396. -. .. .. (24) Tullman, R. H.; Hanzlik, R. P., to be submitted for publication.
_~.
The cytochrome P-450 enzymes are a family of heme-containing proteins involved in the metabolism of a range of substances that possess great diversity in structure. These monooxygenase enzymes catalyze several oxidative reactions that have a common unifying feature: the covalent incorporation of one atom of oxygen into the substrate m ~ l e c u l e . These ~ enzymes are subject to mechanism-based or suicide inhibitionS by numerous substrates as a function of the metabolism of uniquely positioned olefinic,6 acetylenic,' or cyclopropylamine8 moieties. Cyclopropylamines are also suicide inhibitors of mitochondrial monoamine oxidaseg and of plasma amine The mechanism-based inhibition of these distinct enzyme classes by cyclopropylamines has been postulated to be a consequence of a-hydroxylation to form a cyclopropylcarbinolamine, which reacts with enzyme-bound nucleophiles via a highly reactive iminium ion (Figure l : l a 2 3).8*9bWe present evidence here that an alternative mechanism, postulated to involve initial amine nitrogen oxidation to a radical cation, is operative for the suicide inactivation of cytochrome P-450 by cyclopropylamines. We have investigated the in vitro .mixed-function oxidative metabolism of cyclopropylbenzylamine, la, and (1-methylcyclopropyl)benzylamine, lb, by hepatic microsomes and by a reconstituted hepatic cytochrome P-450 system, both isolated from phenobarbital-pretreated rats.I0 Cytochrome P-450 content was determined after metabolism of both compounds directly by spectral assay and indirectly by biochemical assay of enzyme activities. Cytochrome P-450 was determined by its ferrous carbonyl Soret absorbance," and total heme was determined as the pyridine hemochromagen.11*12d-Benzphetamine'O and am-
-
-
(1) Research Fellow of the Alfred P. Sloan Foundation (1981-1983). (2) National Research Service Award Fellow supported by USPHS Training Grant ES07028. (3) Recipient of USPHS Research Career Development Award (ESOOO41) (1 978-1 983). (4) For recent reviews see: (a) Ullrich, V. Top. Curr. Chem. 1979,83,67. (b) Testa, B.; Jenner, P. "Drug Metabolism: Chemical and Biochemical Aspects"; Marcel Dekker: New York, 1976. (c) Goldstein, A.; Aronov, L.; Kalman, S.M. "Principles of Drug Action: The Basis of Pharmacology"; Wiley: New York, 1974; pp 227-300. (d) Ullrich, V. In 'Microsomes and Drug Oxidations"; Pergamon Press: New York, 1977; p 192. (5) For reviews of mechanism-based or suicide enzyme inhibition see: (a) Seiler, N.; Jung, M. J.; Koch-Weser, J. "Enzyme-Activated Irreversible Inhibitors"; Elsevier/North Holland Biomedical Press: Amsterdam, Netherlands, 1978. (b) Walsh, C.; Cromartie, T.; Mariotte, P.; Spencer, R. Methods Enrymol. 1978,53,437. (c) Rando, R. R. Ibid. 1977,46, 28. (d) Abeles, R. H.; Maycock, A. L. Acc. Chem. Res. 1976, 9, 313. (6) (a) Ortiz de Montellano, P. R.; Mico, B. Arch. Biochem. Biophys. 1981, 206, 43 and references therein. (b) Ortiz de Montellano, P.; Beilan, H. S.; Kunze, K. L.; Mico, B. A. J. B i d . Chem. 1981, 256,4395. (7) (a) White, I. N. H. Biochem. J . 1978, 174, 853. (b) Ortiz de Montellano, P. R.; Kunze, K. L. J. B i d . Chem. 1980, 255, 5578. (8) (a) Hanzlik, R. P.; Kishore, V.; Tullman, R. J . Med. Chem. 1979, 22, 759. (b) Hanzlik, R. P.; Tullman, R. H.; Kishore, V. In 'Microsomes, Drug Oxidations, and Chemical Carcinogenesis"; Vol. I.; Coon, M. J., Conney, A. H., Estabrook, R. W., Gelboin, H. V., Gillette, J. R., O'Brien, P. J., Eds.; Academic Press: New York, 1980; Vol. I, p 327. (9) (a) Alston, T. A. Phurmacol. Ther. 1981,12, 1 . (b) Silverman, R. B.; Hoffman, S. J.; Catus, W. B., 111 J . Am. Chem. SOC.1980, 102, 7126. (c) Silverman, R. B.; Hoffman, S.J. Ibid. 1980,102,884. (d) Tipton, K. F. In "Enzymatic Basis of Detoxication"; Vol. 1, Academic Press: New York, 1980; Vol. I, p 355. (10) Guengerich, F. P.; Martin, M. W. Arch. Biochem. Biophys. 1980, 205, 365. (11) Omura, T.; Sato, R. J . Biol. Chem. 1964, 239, 2370. (12) Guengerich, F. P. Biochemistry 1978, 17, 3633.
0002-786318211504-2050%01.25/0 0 1982 American Chemical Society
